How to set a marketing budget from your numbers, not a percentage
Asked for a marketing number, founders copy a 7-to-10-percent-of-revenue rule measured on billion-dollar companies. The percentage is an output, not a decision — set the budget from your growth goal, your cost to win a customer, and your payback window.
By Stacey Tallitsch | July 28, 2026
You have grown the business to $3M, maybe $4M, mostly on work that referred itself. Now something has changed. A banker wants a number for the loan file. A new board member wants a line in the plan. Or the fiscal year simply turned over and, for the first time, someone expects you to commit to a marketing budget instead of spending when a slow month scares you. So you open a browser and type "how much should I spend on marketing." The first several results agree: 7% to 10% of revenue. You do the multiplication, land on a number, and write it into the plan.
That number is fiction. Not because the arithmetic is wrong. Because you solved the wrong equation.
The percentage is an output, not a decision
A marketing budget is not a percentage of revenue. It is an amount of money bought a certain way for a certain return, and the percentage is just what that amount happens to work out to after the fact. When you set the budget by picking a percentage first, you are writing down the answer before you have asked the question. You are deciding to spend $280K next year because a blog told you 7%, not because you know what $280K is supposed to buy.
It gets worse when you look at where that benchmark comes from. The most-quoted figure in every one of those articles traces back to the same place. Per Gartner's 2025 CMO Spend Survey, marketing budgets have flatlined at 7.7% of company revenue. Read one line further and you find the survey polled 402 marketing leaders, the vast majority at companies with more than $1 billion in annual revenue.
Sit with that. The number a founder running a $3M service business copies off the internet is measured on companies a thousand times his size. Those companies carry brand equity he does not have, a category position he does not have, a sales team he does not have, and fixed marketing overhead — agencies, tooling, a department — that eats a chunk of the percentage before a single new customer is pursued. The 7.7% describes the maintenance cost of an enormous machine. It says almost nothing about what a small, hungry business should spend to grow. Copying it is the same mistake as pricing your ads against the market leader who advertises everywhere: you are calibrating to a company that looks nothing like yours.
So throw the percentage out as a starting point. You will use it later, once, as a guardrail. Never as the decision.
Start from the goal and the unit you are buying
The budget serves a goal, so name the goal in units you can count. Not "grow revenue." How many new customers do you need next year, on top of the ones who will come back on their own? A roofing company that wants to add $600K in new work and books an average job of $12K needs roughly 50 new customers it would not have gotten anyway. That is the target the entire budget exists to hit.
Now the two numbers that turn a goal into a budget: what it costs you to win one new customer today, and what one new customer is worth to you over the first year. Most founders under $10M cannot answer either with a straight face. That is not a character flaw. It is the actual first line item.
If you do not know your cost to win a customer, your first marketing dollars do not buy ads. They buy the ability to see. That means a way to tag where inquiries come from and whether they book — a spreadsheet and a disciplined front desk will do it before any software does. Founders who skip this step are the ones who later watch their cost to win each customer climb quarter after quarter and cannot say why, because they never measured it when it was cheap to start. You cannot set a budget blind. Buying sight is the highest-return line on the plan, and it is nearly free.
Size the bet to a payback window, not a revenue share
Here is the number the percentage rule cannot see, and the reason it fails small businesses specifically: payback period.
Two companies do $3M in revenue. The first sells a $9K HVAC install, collects the cash inside 30 days, and knows a satisfied install customer buys maintenance and repairs for a decade. The second sells a $9K annual software contract billed monthly, so it recovers the cost of winning that customer over 11 or 12 months. Same revenue. Same deal size on paper. The first business can afford to spend several times more per customer than the second, and spend it far more aggressively, because it gets its money back before the credit card bill is due. The second has to be patient or it runs out of cash before the paybacks land.
A percentage-of-revenue rule is blind to this. It hands both companies the same 7% and calls it discipline. It is not discipline. It is a coincidence dressed as a rule.
The real constraint on how much you can spend is how long you can wait to get it back, given your cash. Figure out roughly what a new customer is worth to you in year one and how fast that money actually arrives. If you recover your cost to win within your comfortable cash window, you can afford to spend up to that cost, per customer, times the number of customers you set as the goal. That product is your budget. It came from your goal, your economics, and your cash reality — three things a benchmark cannot know about you.
Split the budget before you spend it
One more cut, and it is the one most founders miss. Divide the number into two piles: maintain and bet.
Maintain is the money that defends demand you already earn — keeping your reviews strong, staying visible to the people who already know you, holding the channels that already produce. Cutting this feels like savings and is actually how a quiet business gets quieter. Bet is the money aimed at demand you do not have yet: a new channel, a new audience, a test. Only the bet pile is where experiments belong, and a bet is only a bet if you have decided in advance what result would make you keep spending. That is a discipline in itself, and I have written the full method for running a channel test that returns an actual yes or no rather than a shrug.
Now, and only now, take out the percentage. Divide your derived budget by your revenue and look at the figure. If it lands anywhere from the low single digits to the mid teens, you are in the range real businesses your size actually spend. If it comes out wildly higher or lower, that is not a signal to conform to 7.7%. It is a signal to go back and check your inputs: your goal, your cost to win, your payback. The percentage is a smoke detector, not a thermostat. It tells you to look. It does not set the temperature.
What to do before you close this tab
Do not set a number today. Write down two instead: what it costs you, right now, to win one new customer, and what one new customer is worth to you across the first year. If you can fill both in with real figures, your budget is an afternoon of arithmetic away. If you cannot — and most founders cannot — then you have found your first line item, and it costs almost nothing. Buy the ability to see before you buy anything to be seen with. Everything downstream of that is guessing with a bigger checkbook, which is the one thing a founder your size cannot afford to call a plan.
— Stacey Tallitsch, Stronghold CMO
About the Author
Stacey Tallitsch is the President of Stronghold CMO, a Fractional AI CMO service operating under Talisman Capital, Inc. He is a 30-year tech veteran and the author of 21 books on systems thinking, operator-grade decision-making, and personal sovereignty, with more than 30,000 students across his Udemy course catalog.